A Sports Analogy for Understanding Different Ways to Use AI

18 vocabulary flashcards covering the key concepts, analogies, and design considerations presented in the article about how AI tools can function like steroids, sneakers, or a coach.

Generative AI

A class of artificial intelligence that can create new content (text, images, code, etc.) and is at the center of debates about job displacement versus productivity gains.

Large Language Model (LLM)

A neural-network model trained on vast text data that can generate human-like language, translate, summarize, and answer questions.

Steroid Analogy (AI)

Using AI to get short-term performance boosts that erode human skill over time—e.g., students handing in LLM-written critiques instead of learning to write.

Sneaker Analogy (AI)

Employing AI like performance-enhancing running shoes: momentary productivity gains without long-term de-skilling, such as LLMs for translation or formatting.

Coach Analogy (AI)

Leveraging AI as an intelligent tutor that builds enduring human capability, e.g., adaptive practice questions and feedback for certification study.

De-skilling

The loss or failure to develop human abilities because tasks are offloaded to technology, a concern when AI is used like steroids.

Augmentation

Designing AI to complement and expand human skills rather than replace them, central to the sneaker and coach models.

Spell-check Design Example

A design choice where suggested corrections (rather than automatic fixes) let users learn spelling, shifting the tool from ‘steroid’ to ‘coach.’

Hallucination (AI)

Fabricated or inaccurate content generated by an LLM; must be detected and corrected by users.

Confidence-based Highlighting

A UI feature that flags low-confidence portions of AI output to help users spot hallucinations and maintain cognitive oversight.

Norms for AI Use

Societal or organizational rules that determine when and how AI tools are appropriate, analogous to calculator policies in education and banking.

Calculator Analogy

Historical comparison showing how tool acceptance varies by context—useful for bankers, but detrimental for kids learning arithmetic.

Co-evolution of Skills and Tools

The idea that as AI improves and people adapt, valued human skills shift, similar to changes in emphasis on spelling or long division.

Co-pilot

An AI system designed to work alongside a human, providing assistance while keeping the user in control and engaged.

Nike Carbon-Soled Shoes

Real-world performance sneakers that inspired the ‘sneaker’ analogy—making runners ~4–5% faster without harming long-term ability.

Translation, Reformatting, Annotation

Typical LLM tasks that save time for knowledge workers, exemplifying the sneaker model of AI augmentation.

AI-Powered Tutor

An LLM trained on domain material that generates personalized practice questions and feedback, embodying the coach analogy.

Priority Shift in Skill Development

The reallocation of learning effort as AI automates certain tasks, freeing people to focus on higher-level system design and creativity.